College Admissions Process Can Often Be a Top Source of Stress and Anxiety for Students

October 2018

A “FIT” OVER RANKINGS Why College Engagement Matters More Than Selectivity EXECUTIVE SUMMARY

Based on the Challenge Success survey data of more than 100,000 high school students across the country, we know that the college admissions process can often be a top source of stress and anxiety for students. While many schools, counselors, and parents encourage students to focus on finding the “right fit” college, this advice can be difficult to follow without a better understanding of what “fit” means and what matters most — both for learning and engagement in college — and for life outcomes beyond college.

This paper reviews and synthesizes key research in order to address many of the important questions and concerns we hear from students, parents, and schools about the college admissions process: What do college rankings really measure? Are students who attend more selective colleges better off later in life? What is “fit” and why does it matter? What the research shows:

RANKINGS ARE PROBLEMATIC. Many students and families rely on college rankings published by well-known organizations to define quality. The higher the ranking, the logic goes, the better the college must be and vice versa. We find that many of the metrics used in these rankings are weighted arbitrarily and are not accurate indicators of a college’s quality or positive outcomes for students.

COLLEGE SELECTIVITY IS NOT A RELIABLE PREDICTOR OF STUDENT LEARNING, JOB SATISFACTION, OR WELL-BEING. We explore the research on whether attending a selective college predicts important life outcomes and find no significant relationship between a school’s selectivity and student learning, future job satisfaction, or well-being. We find a modest relationship between financial benefits and attending more selective colleges, and that these benefits apply more to first-generation and other underserved students. We also find that individual student characteristics (such as background, major, ambition) may make more of a difference in terms of post-college outcomes than the institutions themselves.

ENGAGEMENT IN COLLEGE IS MORE IMPORTANT THAN WHERE YOU ATTEND. Colleges that provide ample opportunities for students to deeply engage in learning and campus community may offer the key to positive outcomes after college. For instance, students who participate in internships that allow them to apply what they learn in the classroom to real life settings, students who have mentors in college who encourage them to pursue personal goals, and students who engage in multi-semester projects are more likely to thrive after college.

There is no question that the college admissions process can be stressful. We hope that this paper prompts students and families to examine what college success means to them and to question common assumptions about college selectivity. A good fit is a college where a student will be engaged — in class and out — by what the college has to offer. With over 4,500 colleges in the United States, there are many schools from which to choose. We encourage students and families to look beyond rankings in the college search process, and instead to seek a school where students can participate fully in academic, civic, and social life in order to thrive both during the college years and beyond.

October 2018 INTRODUCTION

Since 2007, we have asked over 100,000 high school experience, students don’t necessarily know what that students, “Right now in your life, what, if anything, means or how to follow that advice. Some students causes you the most stress and why?” The number define “fit” as a match between their SAT scores and one answer is the heavy workload they face. The the average SAT scores of other students at a college. second is usually college admissions. Indeed, most of Many become consumed with the idea that they need the high-performing high schools in the Challenge to be admitted to a “selective” college regardless of fit. Success program list the college admissions process as one of the most common root causes of stress for This focus on selectivity may stem from misconceptions their students. The students and schools identify about college rankings and how college selectivity other stressors as well, such as too many courses and affects both what happens during the college years and extracurricular activities in a day, too little time for life after graduation. In order to help explain what makes deep learning and collaboration, and an over-reliance a good fit and to clear up some of these misconceptions, on testing for assessment, but these are issues that the we conducted an extensive review of the literature on students and schools feel they can ultimately address. college outcomes, such as student learning, well-being, They know they can work to change homework policies job satisfaction, and future income, and we explored and modify schedules and implement more authentic the relationship between these outcomes and rankings assessments, but the college admissions process is and college selectivity. In this paper, we synthesize particularly frustrating to them because it feels like it the current research to address the following three is out of their hands. questions:

1. What do college rankings really measure? This paper addresses some of the concerns of schools, students, and parents who are stressed by the college 2. What is the relationship between college admissions process. We know that students are selectivity and student outcomes? often advised to find the “right fit” college, but in our 3. What is “fit”? Why does it matter?

ABOUT CHALLENGE SUCCESS

Challenge Success is a non-profit organization affiliated with the Stanford University Graduate School of Education. We partner with schools, families, and communities to embrace a broad definition of success and to implement research-based strategies that promote student well-being and engagement with learning.

www.challengesuccess.org

3 What do college rankings measure?

National and international rankings of colleges1 and heading of “selectivity.” In these ranking systems universities play an increasingly large role in the way and in much of the college research, the term “non- students, parents, administrators, and researchers selective” is used for those colleges that admit nearly understand the landscape of higher education. Because every student, for example, community colleges that of the influence of the rankings, many colleges work only require evidence of a high school diploma or hard to conform with the formulae used by the ranking equivalent. The term “selective” typically refers to agencies, particularly the early commercially successful colleges that “select” students to admit from a pool agencies like Barron’s, The Princeton Review, and, of applicants. Depending on the research, “selective” especially, U.S. News and World Report (Hazelkorn, may refer to colleges that admit very high percentages 2015). In fact, the rankings have become so popular, that of applicants as well as those that admit much smaller many people consider selectivity and ranking to be the percentages.2 same thing. The logic is that the more competitive the admissions process, and the more difficult it is to get Let’s take a closer look at how U.S. News and World into a particular college, the better that college is. Report, a particularly prominent ranking agency, determines its rankings for National Universities and Ranking systems and researchers use a more specific Liberals Arts Colleges (they use a slightly different definition of selectivity.Barron’s , for example, sorts formula for “Regional” institutions). Doing so will colleges into tiers based on selectivity as measured provide insight into how ranking lists are created by SAT scores, high school GPA and class rank, and and what they mean. The chart on the following page acceptance rate (Barron’s College Division, 2018). U.S. explains the fifteen metrics used byU.S. News and News and World Report also uses SAT scores, class World Report in 2017,3 though many other ranking rank, and acceptance rate in its formula, all under the agencies use similar metrics (Barron’s College

1 Following U.S. colloquial conventions, we use the terms “college” and “university” interchangeably.

2 In this paper, we do not use the terms “ranking” and “selectivity” interchangeably. When we refer to rankings, we mean where a college is ranked according to a particular ranking system. When we refer to selectivity, we use the more specific definition according to the research we are citing each time.

3 The weights cited and discussed below come from the 2017 U.S. News and World Report website (see https://www.usnews.com/education/ best-colleges/articles/ranking-criteria-and-weights). Note that these are recalculated to provide the exact weight of each metric. In the original rankings, there are seven major ranking areas, each made up of one to five of the listed metrics. For example, “Graduation and Retention Rates” make up 22.5% of the total ranking. 80% of “Graduation and Retention Rates” comes from the graduation rate, while 20% comes from the first-year retention rate. Thus, 18% of the total score (22.5%x 80%) comes from graduation rate, while 4.5% (22.5% x 20%) comes from the first-year retention rate. All percentages are rounded to one decimal place.

4 U.S. NEWS AND WORLD REPORT: “Best Colleges” Ranking Criteria & Weights (2017)

GRADUATION RATE is a measure of what ALUMNI GIVING RATE measures what percentage of students who enroll in a school percentage of graduates contribute graduate within 6 years. A higher graduation financially to the college. Higher is rate is considered better. considered better, as it implies that graduates valued their time at the college enough to donate. REPUTATION AMONG PEERS is calculated by surveying university administrators, then averaging the results. It is a subjective FIRST-YEAR RETENTION is how many measure of prestige. students who enroll at the college stay after their first year. Higher is considered better.

EXPENDITURES PER STUDENT is the amount of money a college spends on is a instruction, research, public service, TOP 10% OF CLASS IN HIGH SCHOOL measure of how many students at the college academic support, student services, were in the top of their high school classes and institutional support (adjusted by GPA. This is an alternative measure of logarithmically). Higher is considered better. student quality, compared to SAT, as some students may not do well on the SAT, but still excel in high school. Higher is considered SAT SCORES are an average across all better. students at the school. ACT scores are recalculated to match the SAT score. FACULTY WITH TERMINAL DEGREES is a measure of the proportion of faculty AVERAGE CLASS SIZE is a measure of how who have PhDs, MDs, or another top-level many students are in each class. Lower is degree, depending on their field. Higher is considered better. considered better.

REPUTATION AMONG GUIDANCE ACCEPTANCE RATE is the number of COUNSELORS, like peer reputation, students who are admitted to the school, is calculated by surveying high school 4 divided by the number who apply. Lower is guidance counselors. It is a subjective considered better. measure of prestige.

FULL-TIME FACULTY RATIO is a measure PROJECTED GRADUATION RATE differs of the proportion of faculty who are employed from actual graduation rate because it is an full-time by the college, usually in positions educated guess as to what percentage of new that combine teaching and research, as students will graduate, rather than a measure opposed to part-time or adjunct faculty, who of previous graduation rates. It takes into usually have only teaching positions. Higher account institutional changes over time, is considered better. including changes in actual graduation rates over the past six years, and projects those changes forward. As with graduation rates, STUDENT-FACULTY RATIO is the number higher is considered better. of students divided by the number of faculty. This correlates closely with class size, as smaller classes necessitate more faculty, FACULTY COMPENSATION is a measure of meaning this ratio gets lower. Lower is faculty salaries, including benefits, adjusted considered better. for regional cost-of-living rates. Higher compensation is considered better.

5 Division, 2018; The Princeton Review, n.d.5). The A closer look at some of these metrics raises even percentages indicate what percent of the total ranking more questions. For instance, considering the list of score comes from each metric. Colleges are ranked expenditures, where do construction expenses and according to the scores derived from this set of metrics. infrastructure costs factor in? Does the alumni giving rate metric encourage colleges to admit students who Rankings may seem like they are objective measures of are more likely to be able to afford to donate after quality because they use complex formulae and present graduation? And why are colleges with more faculty their findings definitively. Ellen Hazelkorn notes in members with higher degrees regarded more favorably? Rankings and the Reshaping of Higher Education that Faculty with terminal degrees may do more research, “differences between institutions are often statistically but they may or may not be experts in teaching or insignificant” (2015, p. 52). Indeed, each of the metrics advising. We won’t go into all of our questions for each listed above seems to be carefully selected, but the metric here, especially since different ranking systems weight each is given in calculating an overall score measure them differently, but the rest of this section is artificially precise. Against what standard are these highlights some of our biggest concerns. weights evaluated? Other ranking systems? Past years’ results? Do they get modified year to year to ensure In the following paragraphs, we discuss a few of the some changes in the rankings? There is no objective metrics that are more heavily weighted. These are the way to decide whether the rankings produced by this major drivers of ordinal differences in the rankings, but selection and weighting of metrics are accurate or the differences are often not particularly meaningful. meaningful. What’s more, some colleges engage in These categories and weights (used here and in other ranking systems) are often unclear and can Ranking systems generally use be manipulated. Moreover, there are many possible measures of quality that are missing entirely, such as data that are easy to gather, not long-term learning outcomes, student happiness on necessarily data that are the campus, graduate satisfaction, and civic engagement. most meaningful. Granted, many of these are hard to measure or compare, but that’s exactly the point: ranking systems generally creative reporting of metrics to present themselves in use data that are easy to gather, not necessarily data the best light (Stevens, 2007). that are the most meaningful.

4 U.S. News and World Report uses the term “guidance counselors,” but some high schools use slightly different terms like “college counselors” or “college advisors.” For the purposes of this metric, these terms can be used interchangeably.

5 The Princeton Review uses some different metrics but overlaps withU.S. News and World Report in many categories, namely: class rank of incoming students, standardized test scores, acceptance rate, class size, student-teacher ratio, full-time faculty ratio, and institutional financial resources. As mentioned above, theBarron’s selectivity index looks specifically at SAT scores, high school class rank and GPA, and acceptance rates. Note that Barron’s does not rank ordinally, and instead groups colleges into broader selectivity categories, from “most selective” to “non-selective.”

6 Graduation Rates and Projected items because there is little transparency to the way Graduation Rates they are calculated. The inner workings of these metrics are mysterious and poorly understood except by a few Graduation rate is the most heavily-weighted category in the know. Whereas figures like graduation rates, SAT and is in a sense counted twice because of the inclusion scores, and acceptance rates are public knowledge, the of a “projected graduation rate” metric (which is survey participants and results that lead to these black intended to show whether schools are improving their box metrics are not identified or reported. graduation rates). These two metrics add up to 25.5% of each school’s score in the rankings. Actual graduation It’s safe to say that most college administrators and rates are readily available from the Integrated high school counselors don’t know the inner workings Postsecondary Educational Data System (IPEDS). A of even a few — much less all — colleges well enough typical highly-ranked private university might have a to accurately evaluate their quality on a year-to-year graduation rate around 95%, while flagship public state basis. A somewhat low-profile college might have just university rates vary widely from state to state (ranging invested in an innovative academic program, but its from over 90% to less than 50% in a few states, with most overall reputation is unlikely to change because most between 65% and 85%).6 However, research suggests college administrators and high school counselors won’t the chance of any particular student graduating in be aware of that program unless they are specifically six years is not related to selectivity once background paying attention to that institution. Reputation, thus, characteristics like family income are controlled for is something of a self-fulfilling metric. University (Heil et. al., 2014). In other words, individual student administrators and high school counselors are likely to characteristics — including family wealth — tend to give good evaluations to schools they have heard more drive student graduation outcomes, as opposed to about, which may be those schools that are already at the schools themselves (or even peer effects at those the top of the rankings (Hazelkorn, 2015). schools). Nevertheless, this metric remains a major factor in rankings, despite being more a reflection of The problem here is that reputation is supposed to student characteristics than of institutional quality. capture exactly these more qualitative shifts in what’s going on in colleges. If a graduation rate dips, or expenditures grow, ranking systems notice and shift Reputation…What’s in the Box? accordingly. If, however, the quality of instruction at a Two of the most important components of the previously college improves, the only metrics that might capture described ranking system are peer reputation and that change are peer and counselor reputation, but guidance counselor reputation (combined to account keeping track of such changes across the entire higher for 22.5% of a school’s score). These are “black box” education landscape (and then turning that into an

6 All IPEDS data are publicly available at https://nces.ed.gov/ipeds/

7 ordinal ranking) is an unreasonable task for already may be more informative. busy administrators and high school counselors.

Other Approaches to Rankings The Problem with Class Size There are a number of different ranking systems Class size makes up 8% of a college’s score and is available these days, each with its own particular calculated by giving schools more points for the weightings and categories. Some rankings have subtle proportion of their classes that are smaller. While differences in metrics without much explanation. Why small classes are often seen as desirable, this metric should rate of alumni giving, for example, be worth is problematic for reasons articulated in How College exactly 5%, while first-year retention is worth 4.5%? Works (Chambliss and Takacs, 2018). When schools What accounts for the 0.5% difference? If you swapped offer more small classes, there is less space for students some of these metrics around, you would get only in those classes, meaning fewer students actually get to slightly different results. take them. Consider, as an extreme example, a school with 200 students where each student takes only one There are, however, some other approaches to rankings class. If there are 9 classes with 2 students each, and that measure colleges according to outcomes like one huge class with the remaining 182 students, the value added to expected income (that is, how much average class size at the school would still be only more money graduates make because they attended 20 students, and a full 90% of those classes would be a specific institution). For example, in its first edition, Brookings ranked three maritime academies, various Rankings are problematic, and technical colleges, and a few schools with specialized business and entrepreneurship programs above high school students would do most of the colleges ranked in the top 100 in more well to broaden their gaze. popular ranking systems (Rothwell, 2015). Many of the highest-ranking schools in this ranking system considered “small” classes. However, less than 20% of are non-selective colleges that accept all or nearly all the students at the school would be enrolled in a small applicants, or selective colleges that admit over 70% class. of their applicants. If your goal in attending college is simply to maximize your income, attending one of these Rather than measuring the proportions of class sizes, easier-to-get-into, less expensive programs at a lesser- emphasis should be put on the number of students who known school is worth considering. Other ranking have the opportunity to enroll in and complete smaller systems look at criteria such as the number of low- courses. Of course, this is a much harder figure to income and first-generation students enrolled (such determine, which echoes a larger issue with rankings: as Washington Monthly; see Carey, 2017) or surveys of they are based on institution-level data that are easy student satisfaction (Niche, for example). Because of to collect, and not necessarily student-level data, which these different metrics, the colleges that end up at or

8 near the top of these rankings differ from those at the top of the more popular ranking systems.

Rankings in Summary

While any ranking system has its own internal rhyme and reason, there is little objective basis for any particular set of metrics and weightings. In the words of one economist, “There is no such thing as a single index of quality that uniformly affects all possible outcomes [for students in college]” (Long, 2008). It seems impossible to reduce the quality of an institution to a single number, and then to sort schools by that number in any kind of rigorous, accurate way, especially given the wide range of goals students have as they enter college, as well as the tremendous variability in student backgrounds.

Traditional college rankings measure a set of factors that are weighted arbitrarily, drawn from data that are most easily quantifiable and comparable, sometimes poorly documented, and not always relevant to undergraduate education. In short, rankings are problematic, and high school students would do well to broaden their gaze. There are better ways to choose a college than to rely on rankings and, especially, selectivity as the main criteria, as we explain below.

9 What IS THE relationship between college selectivity & student outcomes?

Much of the research on college outcomes looks intellectually during their time in college. The results primarily at financial results from college attendance, here are clear and have been consistent for decades. including short-term employment and long-term In the most recent edition of their literature review on earnings. This is an important outcome, both for college outcomes, Mayhew et al. (2016) write: individual students and their families, and for society. Little evidence suggests that selectivity However, a lot more happens in college than career is related to measures of students’ self- preparation. For example, Matthew Mayhew and his reported gains in learning, let alone verbal, co-authors (2016) divide their 600-plus-page review of quantitative, or subject-matter competence over 1,800 peer-reviewed research studies on the effects measured by standardized tests. This of college into eight major categories, ranging from finding, which has been consistent over growth in basic quantitative and verbal skills to moral the past 40 to 50 years, has implications development to changes in social attitudes to quality for college choice, family finances, and of life impacts. In short, college affects students in a public policy, particularly as students, variety of ways: economically, socially, psychologically, their families, and policymakers deal with and intellectually. the differential allocation of resources of publicly supported institutions. In the following paragraphs, we’ll examine a few Accounting for student background outcomes that tend to be of particular concern characteristics, the weight of evidence to families: student learning, job satisfaction and simply does not support students’ or general well-being, and financial outcomes. And we’ll policymakers’ beliefs that a selective explore the relationship between these outcomes and admissions process enhances student selectivity. Do the most selective colleges lead to the learning. (p. 96) best outcomes in terms of learning, well-being, and future income? What does correlate with student learning? Time spent studying (Arum and Roska, 2011). This is true regardless of institution and for all kinds of students. College as a Place of Learning In other words, a student who studies hard at a non- Students and their families hope that students will selective school is very likely to learn, while a student learn new skills, acquire new knowledge, and grow who slacks off at a selective one is less likely to learn.7

7 In addition to studies showing no difference in student learning at more selective schools, two studies have found that civic values and engagement — for example, likelihood of voting or doing community service — decrease when attending a selective institution (Astin and Antonio, 2004; Lott, 2013).

10 What are the unstated implications of these findings? and physical, using ten survey questions to determine There are several, but most important for college-going in which categories graduates demonstrate they applicants and their families is that they would be well- are “thriving,” “struggling,” or “suffering.” They then served by casting a wider net in their college searches. compared how many graduates report thriving in one, There is no evidence that students’ learning will suffer two, three, four, or all five categories. They found no relationship, in most cases, between the type,8 location, There is no evidence that students’ or selectivity of college and well-being after graduation. Similarly, workplace engagement, as measured by a learning will suffer for attending a less 12-question “Engagement Index” that Gallup uses in selective college. other studies as well, did not differ between graduates of more and less selective colleges. for attending a less selective college, and they may well find a better fit in other ways. In particular, students are While Gallup-Purdue acknowledges that college best served by attending institutions where they are decisions — where to apply, where to attend — are not motivated to invest meaningful time in their studies. easy, they argue that, We’ll return to this in the final section of the paper. The data presented in this report suggest …that the answers lie in thinking about College as Preparation for Job things that are more lasting than selectivity Satisfaction and General Well-Being of an institution or any of the traditional measures of college. Instead, the answers Since 2014, Gallup-Purdue has conducted a survey of may lie in what students are doing in job satisfaction and general well-being of graduates college and how they are experiencing it. from a wide variety of colleges. The 2014 survey results (Gallup Inc., 2014, p. 5) included data from 1,557 associate degree holders and 29,560 bachelor’s degree holders. Rather than Gallup-Purdue’s findings echo other prior research on focusing narrowly on income, this survey looked more job satisfaction and well-being. For instance, ten years holistically at how college graduates feel about their after graduation, former students of selective colleges jobs and careers. report lower job satisfaction than students from less selective colleges; students from more selective The 2014 Gallup-Purdue annual report found no colleges are more likely to feel underpaid (Liu et al., relationship between college selectivity and both 2010). workplace engagement and general well-being (Gallup Inc., 2014). Gallup-Purdue measured well-being along five dimensions: purpose, social, financial, community,

8 The only exception is private, for-profit institutions, the graduates of which show significantly lower well-being.

11 College as an Economic Opportunity selective college.

If selectivity does not predict learning, well-being, or In a recent study that compares what Barron’s calls job satisfaction, does it predict economic outcomes? the “most selective” schools to non-selective ones, There are many different ways researchers define Witteveen and Attewell (2017) find a similar relationship and measure selectivity, which can lead to differing to that found by Zhang. Controlling for student major conclusions about economic outcomes. This makes it and enrollment in post-graduate studies (which can problematic to compare findings. In the paragraphs increase income), they find a 21% difference in earnings below, we’ll present various findings — using, for the between the most selective and non-selective schools sake of accuracy, the language from the original studies (again, schools that admit almost any student). as much as possible — and we’ll highlight the broader takeaways. Additional studies look more narrowly at average SAT scores of an entire class of incoming students, rather There is some evidence to suggest that institutional than using Barron’s broader selectivity index. This selectivity is associated with long-term financial allows us to better understand the differences between outcomes. In one seminal study, Zhang (2008, 2012) similarly selective schools. Seki (2014) finds a 3% finds that the graduates of “high quality” (Zhang’s increase in wages associated with a 118-point difference term) institutions have salaries approximately 6 to in average SAT scores of incoming students. Chalak and 8% higher directly out of college than graduates from Kim (2017) find that a 100-point difference in average what he calls “low quality” institutions (which refers to SAT scores of incoming students is associated with, “non-selective” colleges that admit almost any student at most, a 4.8% increase in earnings. These differences who applies), with that percentage rising to 16 to 19% are statistically significant, but are small compared to a decade after college.9 While the average earnings of the differences between the most selective and non- a graduate of these “high quality” colleges are higher selective schools, and again are quite small compared than those from non-selective colleges, Zhang’s data to differences among graduates of individual schools. demonstrate much larger variation within institutions than between institutions. That is, the difference Is it the College, or is it the Student? between the lowest earners and the highest earners from Zhang’s “high quality” colleges is much greater Though the studies above show economic benefits than the difference between the average graduate from to attending selective institutions, researchers have that college and the average graduate from a non- not found a simple and commonly agreed upon way

9 Zhang divides three groups of over 5,000 students (from three different sample years) based on theBarron’s selectivity index of the colleges they attend. Barron’s calculates selectivity using SAT scores, high school GPA and class rank, and acceptance rates, and has several tiers, which Zhang sorts into six categories: high quality public, high quality private, medium quality public, medium quality private, low quality public, and low quality private. The top group of “high quality” public and private institutions contains approximately 100 colleges.

12 to account for the fact that students self-select into the students who attended the more selective schools selective colleges. Because students aren’t randomly outperformed their matched peers who chose to attend assigned to colleges, it’s difficult to determine whether less selective schools. the benefits of attending a more selective college result from attending that college, or rather reflect the kinds Dale and Krueger found, instead, that selectivity had of students who attend those colleges in the first place. no effect on long-term earnings. The same students Is the value added by the college or by the individual who were admitted into the selective schools but who student? chose to attend less selective schools did just as well financially later on in life. One important caveat: Dale Dale and Krueger, in a study completed first in 1999 and and Krueger, among others, have found small benefits revisited twice since with similar findings, including from attending more selective institutions for students most recently in 2014, investigated exactly this question who are the first in their generation to attend college (Dale and Krueger, 2002, 2014). They used the College and for traditionally underserved students.10 For and Beyond dataset, which surveyed students from most students attending these selective universities, 34 colleges, most of which are in the top three tiers of however, the research shows no difference in financial Barron’s selectivity index (roughly 200 institutions), outcomes. Note that Dale and Krueger’s findings as well as a small set of non-selective schools. The average SAT scores of incoming students at colleges in Differences are driven by individual the sample range from the 800s to nearly 1400. students, not by the quality of the student body as a whole or the Dale and Krueger formed matching groups of students who were accepted to similar (and often the same) institutions themselves. selective institutions. Within these matched groups, which in the latest analysis of the data included cannot be generalized beyond the selective schools in 14,238 students, some students attended the more their study, but their research suggests that we should selective schools, while others turned down admission consider the possibility that even in studies that find offers from selective schools and opted to attend less differences in economic outcomes between selective selective alternatives — for any number of reasons, and non-selective colleges, these differences are driven including financial concerns, or because they wanted by individual students, not by the quality of the student to stay closer to home. If selectivity is truly the driving body as a whole or the institutions themselves. force behind better wages, we would expect to find that

10 Roughly 1/10 of a standard deviation in income outcomes. Similarly, Chetty et al. (2017) find that college can be a major source of social mobility for low-income students, and that the most selective institutions in fact do not do a good job of enrolling the students who would benefit most from attending. For a comprehensive discussion of underrepresented populations in higher education, including admissions and outcomes, see Bowen and Bok’s (2000) seminal book, The Shape of the River.

13 Mayhew et al. (2016) agree with this conclusion in their of America’s elite.11 Does attending a college that offers synthesis of this economic outcome research. They access to large percentages of students from upper and write: upper-middle class backgrounds provide a “network effect” that results in better job prospects and, thus, The estimated earnings effects of having an increase in earnings over time? This is a difficult a bachelor’s degree from a selective question to answer and one not directly addressed in institution may be inflated. Studies current research. The research we’ve discussed suggests generally did not control for measures of that, even if network effects may be a driving force individual ambition and other unobserved characteristics, which may influence students’ pathways into selective colleges While some employers might and universities, as well as their subsequent check the name on your college productivity and earnings. (p. 437) transcript, most care far more The benefits of attending the most selective colleges about your track record in the seen in some of the studies above are challenged by field, basic communication and researchers such as Dale and Krueger, who use this problem-solving skills, and the specific methodology to account for individual student attitude and work ethic you differences. Those benefits, furthermore, are most bring to the table. pronounced when comparing selective colleges to non- selective ones, not when comparing selective colleges to each other. between differences in earnings outcomes between colleges, that effect is much smaller than differences Research on the economic outcomes of college tends driven by experiences within individual colleges. not to investigate the mechanisms for why students do or do not make more money depending on the college Selectivity in Summary they attend. The research on learning and selectivity shows little to no relationship between the two, so we Job satisfaction, general well-being, learning, and cannot point to academic differences per se to explain income are all important outcomes of college. Research if and how attending a selective institution may result shows no relationship between selectivity and learning, in higher earnings. One popular belief is that going job satisfaction, or general well-being. There may be, to a highly selective school is not necessarily about at best, modest financial benefits to be gained from learning more, it’s about gaining access to a network attending more selective colleges, but the research

11 Selective admissions historically arose as a way to reinforce social class for the early-20th century New England elite (Karabel, 2006; Wechsler, 2014), and today, the highest ranked institutions still enroll a disproportionately large percentage of students from upper and upper-middle class backgrounds (Leonhardt, 2018).

14 is mixed and much more nuanced when it comes to the long-term financial benefits of attending a more selective school, once student characteristics are taken into account. Notably, there is little variability among selective colleges, and within-college effects (such as academic major and individual student effort) far outweigh between-college effects (Mayhew et al., 2016).

Most rankings tell you primarily how famous a school is. As the studies above suggest, selectivity and prestige are not a one-way ticket to financial success, let alone a happy and fulfilling life. While some employers might check the name on your college transcript, most care far more about your track record in the field, basic communication and problem-solving skills, and the attitude and work ethic you bring to the table (NACE, 2017).

15 What matters? Engagement.

We’ve been focusing on the differences between increased subject-matter competence and general colleges, but we continue to see that variability within knowledge (Berger, 2002; Rocconi, 2011). Engagement colleges is more important for a range of student with coursework also increases curiosity, creativity, outcomes. The reason for this is that there is so much and initiative (Pike et al., 2012). Unsurprisingly, variability in students themselves. In this section we the more students study, and the more intentional shift our focus away from college-level outcomes and their use of study skills, the better students do in towards students and their individual outcomes. their courses (Arum and Roska, 2011; Johnson and Kuennen, 2004; Stump et al., 2011). Outside of the In our reading of the research on student outcomes classroom, engagement in campus community, clubs, — learning, financial, and otherwise — this theme arises: the students who benefit the most from college are those who are most engaged in their academics The students who benefit the most and campus communities, taking advantage of the from college are those who are opportunities and resources their particular institution most engaged in their academics provides. Engagement is the key. and campus communities, taking advantage of the opportunities Mayhew et al. summarize the research on engagement and resources their particular in higher education by saying, “It appears that what institution provides. students do in college is far more important than the Engagement is the key. type of institution they attend” (Mayhew et al., 2016, p.38; See also: Astin, 1997; Bruni, 2016; Pascarella and Terenzini, 1991, 2005). They add, “Students learn and extracurricular activities is associated with a range more when actively engaged in the learning process” of positive outcomes, from cognitive development (Mayhew et al., 2016, p. 60). to identity formation and psychosocial development to moral and ethical development to persistence to Specifically, effort put into coursework results in improved grades (Trowler and Trowler, 2010).12

12In the previous section we discussed the importance of controlling for background characteristics when considering the long-term financial effects of attending particular colleges. While many of the studies in this section on engagement explicitly measure and make methodological or analytical adjustments based on student background characteristics, it’s worth noting that the unit of analysis in most of these studies is individual students, and not institutions. That is, these studies ask how student behavior and activity in college impact certain knowledge, skills, abilities, or characteristics, not how the college attended impacts the students’ outcomes. As a result, background characteristics are less relevant to these research questions and results, and furthermore are built into the sampling and methods.

16 The Gallup-Purdue studies mentioned above show Chambliss and Takacs (2018) similarly find that the a strong connection between certain forms of relationships students cultivate in college — especially engagement in college and future job satisfaction and with professors as mentors — are highly related to their well-being. They report six key college experiences learning, satisfaction with the college experience, and that have an impact on how fulfilled employees feel at post-graduate success. work and whether they are thriving in life after college: Considering that these college experiences are tied 1. TAKING A COURSE with a professor who makes to future workplace engagement as well as health and learning exciting. thriving, and that these experiences take place in a wide variety of colleges, regardless of size, location, or 2. WORKING WITH PROFESSORS who care about selectivity, the implications seem clear: A good “fit” is students personally. a college where a student will be engaged — in class and out — by what the college has to offer. 3. FINDING A MENTOR who encourages students to pursue personal goals. For some students, deciding where to go to college may 4. WORKING ON A PROJECT across several depend on academic factors, such as access to cutting- semesters. edge researchers in a beloved field with opportunities to be involved in graduate-level work, or a well- 5. PARTICIPATING IN AN INTERNSHIP that established professional preparation program. For applies classroom learning. others, the decision might be influenced by location: a college close to home, or far away, in a small town or in 6. BEING ACTIVE IN EXTRACURRICULAR a big city. Or it might have to do with extracurricular ACTIVITIES.

(Gallup Inc., 2014) A good “fit” is a college where a student will be engaged — in class And yet, as important as these experiences seem to be, and out — by what the college very few college graduates say that they experienced them. Only 14% of graduates strongly agreed they were has to offer. supported by professors who cared, made them excited about learning, and encouraged their dreams. Only 6% programming: a robust student activities program, of graduates strongly agreed they had a meaningful intramural sports, or the arts. The decision might internship or job, worked on a long-term project, and include cultural opportunities: a university with a were actively involved in extra-curricular activities. strong international languages program, a culturally And only 3% experienced all six factors (Gallup Inc., diverse student body, and a track record of successful 2014). study-abroad exchanges. The presence or absence of a

17 Greek system or a large Division 1 athletics program might be important factors for some students. Perhaps religious denomination, institutional size, or comprehensive support for those with learning differences may sway the decision. And of course, financial aid and cost are key factors as well.

This is hardly an exhaustive set of criteria, but it suggests that students should look at much more meaningful characteristics than selectivity and rankings when thinking about where they want to go to college. Rather than choosing a school based primarily on a flawed scoring system, students should ask whether they will be engaged at the college in ways that will allow them to form strong relationships with professors and mentors, apply their learning via internships and long-term projects, and find a sense of community.

18 CONCLUSION

There are over 4,500 accredited degree-granting attends. colleges in the United States, allowing students a wide variety of choices for their education. Attempting There is no question that the college admissions to disentangle the different variables, inputs, and process can be stressful. Low acceptance rates and outcomes is not easy, and though this paper summarizes sensational stories of students applying to 20 or more several high-quality studies that attempt to account for schools belie the reality that anyone who wants to — a number of variables — including some difficult-to- who is willing to spend the time and make the financial measure characteristics of individual students — more investment (more affordable in the case of most research needs to be done to help students and families community colleges) — can go to college in the United make wise college choices. Consider this: The top- States. The reality is, higher education in the United ranked 5% of colleges — according to rankings that are, States is uniquely, exceptionally strong (Labaree, as we have discussed, problematic — include over 200 2017). We hope that this paper encourages students institutions. Most college applicants and their families and families to look beyond rankings and selectivity cannot name 200 colleges, and yet the differences in the college search process, and instead seek a good between the top of the list and the 200th on the list — fit, a school where students can engage and participate in terms of all the outcomes discussed in this paper — fully in academic and social life in order to thrive both are minimal. Regardless of whether a student attends a during the college years and beyond. college ranked in the top 5% or one ranked much lower, the research strongly suggests that engagement in college, how a student spends his or her time, matters much more in the long run than the college a student

ACKNOWLEDGEMENTS

We are especially grateful to Alice Kleeman, Gabrielle McColgan, Dr. Richard Rende, Dr. Ethan Ris, and Alyson Tom for their thoughtful feedback on early drafts of this paper.

Arum, R., & Roksa, J. (2011). Academically Adrift: Limited Learning on College Campuses. Chicago, IL: University of Chicago Press.

Astin, A. W. (1997). What Matters in College?: Four Critical Years Revisited. San Francisco, CA: Jossey-Bass.

Astin, H. S., & Antonio, A. L. (2004). The impact of college on character development. New Directions for Institutional Research, 2004(122), 55–64.

Barron’s College Division. (2018). Profiles of American Colleges 2019. Barron’s Educational Series.

Berger, J. B. (2002). The influence of the organizational structures of colleges and universities on college student learning. Peabody Journal of Education, 77(3), 40–59.

Bowen, W. G. and Bok, D. (2000). The Shape of the River. Princeton, NJ: Princeton University Press.

Bruni, F. (2016). Where You Go Is Not Who You’ll Be: An Antidote to the College Admissions Mania. New York, NY: Grand Central Publishing.

Carey, K. (2017, August 27). A different kind of college ranking.Washington Monthly, September/October 2017. Retrieved from https://washingtonmonthly.com/magazine/septemberoctober-2017/introduction-a-different- kind-of-college-ranking-9/

Chalak, K., & Kim, D. (2017). Measurement Error Without Exclusion: The Returns to College Selectivity and Characteristics. Rochester, NY: Social Science Research Network.

Chambliss, D. F., & Takacs, C. G. (2018). How College Works (Reprint edition). Harvard University Press.

Chetty, R., Friedman, J. N., Saez, E., Turner, N., & Yagan, D. (2017). Mobility Report Cards: The Role of Colleges in Intergenerational Mobility. National Bureau of Economic Research.

Dale, S. B., & Krueger, A. B. (2002). Estimating the payoff to attending a more selective college: An application of selection on observables and unobservables. The Quarterly Journal of Economics, 117(4), 1491–1527.

Dale, S. B., & Krueger, A. B. (2014). Estimating the Effects of College Characteristics over the Career Using Administrative Earnings Data. Journal of Human Resources, 49(2), 323–358.

Gallup Inc. (2014). Gallup-Purdue Index Report 2014. Retrieved from http://news.gallup.com/reports/197141/gallup- purdue-index-report-2014.aspx

20 Hazelkorn, E. (2015). Rankings and the reshaping of higher education: the battle for world-class excellence (2nd edition.). Palgrave Macmillan UK.

Heil, S., Reisel, L., & Attewell, P. (2014). College Selectivity and Degree Completion. American Educational Research Journal.

Integrated Postsecondary Education Data System. U.S. Department of Education. Institute of Education Sciences, National Center for Education Statistics. Retrived August 7, 2018 from https://nces.ed.gov/ipeds/

Johnson, M., & Kuennen, E. (2004). Delaying developmental mathematics: The characteristics and costs. Journal of Developmental Education, 28(2), 24.

Karabel, J. (2006). The Chosen: The Hidden History of Admission and Exclusion at Harvard, Yale, and Princeton (Reprint edition). Boston, MA: Mariner Books.

Labaree, D. F. (2017). A Perfect Mess: The Unlikely Ascendancy of American Higher Education. Chicago, IL: University of Chicago Press.

Leonhardt, D. (2018, January 20). America’s Great Working-Class Colleges. The New York Times. Retrieved from https://www.nytimes.com/2017/01/18/opinion/sunday/americas-great-working-class-colleges.html

Liu, X., Thomas, S., & Zhang, L. (2010). College quality, earnings, and job satisfaction: Evidence from recent college graduates. Journal of Labor Research, 31(2), 183–201.

Long, M. C. (2008). College quality and early adult outcomes. Economics of Education Review, 27(5), 588–602.

Lott II, J. L. (2013). Predictors of civic values: Understanding student-level and institutional-level effects.Journal of College Student Development, 54(1), 1–16.

Mayhew, M. J., Rockenbach, A. N., Bowman, N. A., Seifert, T. A. D., Wolniak, G. C., Pascarella, E. T., & Terenzini, P. T. (2016). How College Affects Students: 21st Century Evidence that Higher Education Works (1 edition). San Francisco, CA: Jossey-Bass.

Niche. (n.d.) Methodology for Niche College Rankings. Retrieved August 7, 2018 from https://www.niche.com/ colleges/rankings/methodology/

National Association of Colleges and Employers. (2017). Job Outlook 2017.

Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students: findings and insights from twenty years of research. San Francisco, CA: Jossey-Bass.

Pascarella, E. T., & Terenzini, P. T. (2005). How college affects students (Vol. 2). San Francisco, CA: Jossey-Bass.

21 Pike, G. R., Smart, J. C., & Ethington, C. A. (2012). The mediating effects of student engagement on the relationships between academic disciplines and learning outcomes: An extension of Holland’s theory. Research in Higher Education, 53(5), 550–575.

Rocconi, L. M. (2011). The impact of learning communities on first year students’ growth and development in college. Research in Higher Education, 52(2), 178–193.

Rothwell, J. (2015, October 29). Using earnings data to rank colleges: A value-added approach updated with College Scorecard data. Brookings Institution. Retrieved from https://www.brookings.edu/research/using-earnings-data- to-rank-colleges-a-value-added-approach-updated-with-college-scorecard-data/

Seki, M. (2014). Heterogeneous Returns to College Selectivity and the Value of Graduate Degree Attainment (SSRN Scholarly Paper No. ID 2520978). Rochester, NY: Social Science Research Network.

Stevens, M. L. (2007). Creating a Class: College Admissions and the Education of Elites. Harvard University Press.

Stump, G. S., Hilpert, J. C., Husman, J., Chung, W., & Kim, W. (2011). Collaborative learning in engineering students: Gender and achievement. Journal of Engineering Education, 100(3), 475–497.

The Princeton Review. (n.d.). The Princeton Review’s College Ratings. Retrieved August 7, 2018 from https://www. princetonreview.com/college-rankings/college-ratings

Trowler, P., & Trowler, V. (2010). Student engagement evidence summary. The Higher Education Academy.

U.S. News and World Report. (n.d.) Best Colleges Ranking Criteria and Weights. Retrieved August 7, 2018 from https:// www.usnews.com/education/best-colleges/articles/ranking-criteria-and-weights

Wechsler, H. S. (2014). The qualified student: A history of selective college admission in America. Transaction Publishers.

Witteveen, D., & Attewell, P. (2017). The earnings payoff from attending a selective college.Social Science Research, 66, 154–169.

Zhang, L. (2008). The way to wealth and the way to leisure: The impact of college education on graduates’ earnings and hours of work. Research in Higher Education, 49(3), 199–213.

Zhang, L. (2012). Does quality pay? Routledge.